Planning

Summary Timeline

More details below.

Version	Data	Released	Brief Description
2.0	Spitzer	Spring 2019	Translation of IDL version
2.1	Spitzer+Akari	Apr 2021	Extension of 2.0 to include Akari data
2.2	ISO	Fall 2021	Adaptation of 2.0 to model ISO data
2.3	JWST Sim	May 2022	Adaptation of 2.1/2.2 to JWST simulated data
2.4	JWST	Jan 2023	Adaptation of 2.1/2.2 to JWST based on ERS data
2.5	JWST	May 2023	Final ERS delivery

Note: 2.1/2.2 version numbers may change depending on exact development effort

All pull requests will be reviewed by at least one (and hopefully) two of the repository maintainers (currently Karl Gordon, JD Smith, and Els Peeters).
IDL PAHFIT versions are v1.x. Python PAHFIT versions begin with v2.x.
The original PAHFIT and PAHFIT v2.0 publications should be referenced when using the python PAHFIT code at a minimum.

Science model
- Dust features: PAHs, Silicates, Ices, Dust attenuation
- Gas Features: Ionized, Atomic, Molecular (H_2)
- includes intrinsic widths/profiles
Instrument model
- spectral resolutions/profiles
  - allow for variations with wavelength, combinations of instruments, etc.
Fitting
- abstract the choice of different fitters
  - chisqr minimum optimizers
  - MCMC optimizers (e.g., emcee)
  - grid search?
- model constraint packs
  - vary with data quality and source type
  - at least: list of and min/max for feature properties, including ratios between features
  - ideal: full prior specifications for more complicated than min/max flat constraints, parameter ties

Updates:
- Translation of IDL version to Python (with bugfixes)
- Update PAHFIT to use astropy.modeling framework
- Set-up the modular approach for future additions and contributions
Use different fitters
Applicable to: Spitzer data
Object focus: galaxy spectra
Lead by: Karl/JD/Thomas
Developed fully open
- Official version released when it works
- Development versions available, but not advertised until release
Possible training sets / tests (IDL vs Python):
- Spitzer data on galaxies (low and high R): Karin?
- Spitzer data on galaxies of broad types (low R): Henrik?
- Spitzer data on HII regions in M33/M83 (low and high R): Els/Alexandros
Timeline: Spring 2019
Publication:
- "lightweight" publication with all contributors at that point
- could focus on IDL vs python testing and introduction python package
- possible refereed journals: JOSS (https://joss.theoj.org/) or PASP

Updates:
- model extended to include Akari data (shorter wavelengths to 2.5µm)
- Include ices, additional dust features and lines
- Updated attenuation model
Applicable to: Akari+Spitzer data
Object focus: galaxy spectra
Lead by: Thomas Lai and JD
Not advertised until paper released
Training Set / Test:
- Akari + Spitzer data on galaxies: Thomas & JD
- Larger suite of Spitzer data on galaxies (including ices): Henrik?
Timeline: Fall/Winter 2019
Publication: Paper 2 of Thomas’ PhD

Updates:
- Modify so it’s applicable to ISO-SWS wavelength and resolution
- Add continuum and PAH feature options
Applicable to: ISO-SWS
Object focus: PDR spectra (1st priority) & circumstellar spectra (2nd priority)
- can also use Spitzer spectra to get more PDR/circumstellar examples (at lower resolution)
Lead by: Els/Alexandros
Developed fully open
- Official version released when it works
- Development versions available, but not advertised until released
Training Set / Test:
- ISO-SWS data of HII Regions, PDRs, evolved stars(?) & 3 galaxies: Els/Alexandros
Timeline: Fall/Winter 2019
Publication: Yes.

Updates:
- Adapt existing model and instrument packs to simulated and real JWST NIRSpec and MIRI data
Applicable to: JWST NIRSpec and MIRI spectroscopy
Object focus: PDR and galaxy spectra
Lead by: All
Developed fully open
- Official version released when it works
- Development versions available, but not advertised till released
- Version 2.3 based on simulated data before launch
- Version 2.4 based on real data launch + 6 months (or so)
Training Set / Test:
- Simulated JWST data
- ERS(/GTO?) JWST data
Timeline: Launch and launch + 0.5-1 year?
Publications:
- PDR-ERS team using ERS data
- others for galaxy datasets
- others for any data

it is critical to have publicly available sets of data for training and validation of new models.
use them as part of jypter notebooks, tutorials for documentation, automated testing, etc.
Spitzer
- Galaxies: JD/Henrik?
- MW objects: Els?
Spitzer+AKARI
- Broad Galaxy sample: Thomas/JD
ISO
- MW objects: Els?
- Galaxies: Henrik?
JWST
- ERS PDR program: Karl/Els/Alexandros/...
- Resolved/unresolved galaxy spectra: JD/Karin/Henrik/...